1 ISPA 2008

Current ILS Operations and Lessons for the Future

Gerhard E. Berz, EUROCONTROLMarcel Amherd, swiss49 ag

Juan Polymeris, SWISS International Air Lines

International Symposium on Precision Approach and Performance Based NavigationBonn, Germany, 15 October 2008

2 ISPA 2008

Content

Brief Introduction to ILS Sustainment (Localizer)

Operational Flight Data Analysis using the EMS

ILS Operations: Review of Approach PhasesInitial Approach – Entering ILS Coverage – ILS Capture & Tracking

Lessons for the Future

3 ISPA 2008

Interference Mechanisms of Two Frequency SystemsReflection and / or Diffraction from Static and / or Dynamic Objects

Course CourseEffect: Course scalloping

Clearance CourseEffect: Course bend

Clearance ClearanceEffect: Scalloping & possible false axis

MitigationSuper Wide Aperture Arrays

ILS Localizer Challenges & Solutions

Mitigation

Restricting Radiation

Often Results in States Needing to File Deviation from ICAO Annex 10 SARPs

4 ISPA 2008

Raised Coverage Localizer Concept

Limit Clearance Radiation outside ±15 DegreesStrong roll-off at 15°, but soft “shoulders” out to 35° to mask course sidelobesClearance Peak shifts inward from 12 to 7°Mitigates typical “Hangar Reflectors” often found at 12 to 15°

ConsequenceDifficult / impossible to meet minimum field strength at 2000 ft HAT and 35°However: low & far out coverage corners have limited operational significanceSARPs change: permit raising lower boundary up to defined maximum

BenefitsMore stable guidance signals where it matters (centerline and capture sectors)No false axis in ±35° sectors, clean full scale deviations

GoalEnable ILS manufacturers and ANSP to implement ILS LOC that meets challenges by design (instead of accepting deviation from Annex 10)

5 ISPA 2008

Current Status

One Localizer meeting proposed requirements in operationZRH ILS16 Cat III certified since November 2007

ICAO NSP finalizing changes in coverage requirementsPossible agreement at Fall 2008 meetingLower coverage boundary height function of operational need (vertical use)

6 ISPA 2008

Safety Advisory Board Meeting

Printer

ACARS

(Aircraft Communication andReporting System)

up to 2500parameters

MaintenanceBase

SWISS Datawarehouse(EMS System)

Data Analysis via Internet (secure)

FDR

CVR

Troubleshooting, Eng. Trend,Systems Maintenance

Flight Safety, Ops Engineering,Fuel Advisory, Performance, Planning

InternetMaintenance

Groundstation

ParameterDatabase

Data Acquisition +Management

Flight Data Recording System ADAS & EMSAircraft Data Acquisition System & Event Measurement System

7 ISPA 2008

Supporting Operational Studies

Check for Approach & IDENTWhat the pilots do with ILS outside of coverage…IDENT inside coverage and workload

ILS Intercept and CaptureFalse capture issue and coverage

ILS Tracking and Dynamic Multipath

8 ISPA 2008

Tuning of ILS (First Pilot Interest)

First tuning usually in En-Route Phase

IDENT can be received well over 100NM out

Pilots understand & expect that needles will be unreliable

Desired feature for some (‘needles alive”)

IDENT check is normally done during Check for Approach

Around FL100USUALLY well outside of coverage

9 ISPA 2008

Altitude 12000ft- (Approach Brief)- Approach Check- IDENT Check

Radar Vectoring

10 ISPA 2008

IDENT – What’s it for?

Vital safety function especially at uncontrolled airfieldsNo IDENT = ILS signal shall not be used (ICAO Safety Campaign)Confirmation of correct tuning of an operational facility

Report “Established on Localizer” enables hand-back of NAV responsibility to Pilot and final descent on glidepath (FAP)

Good practice not to arm approach mode without IDENTThis “normal ops” function took some time to identify!

In THEORY, Pilots would have to wait with IDENT until within coverage

Practice: Pilots want to get IDENT out of the way as early as possible to reduce high workload flight phase at localizer intercept

11 ISPA 2008

IDENT Study Approach

Construct Worst-Case IDENT ScenarioIf it had to be done within coverageDetermine if IDENT can be completed given workload

Fleet ConsiderationsFocus on large air line operations

IDENT via PFD DisplayRegional, Business and GA Fleet less critical:

Audio IDENT comes earlier than PFDHigher dynamics correspond to aircraft capabilities (EMS Data, London City)IDENT often also available via DME (ILS associated DME)

12 ISPA 2008

Worst Case Geometries & Timing

Downwind Approach: close to LOC, many sidelobesStraight-In Approach: main signal lobe, time to acquire signalMost difficult are T-Approaches

least time b/w coverage & LOC intercept

EMS shows distribution at 3 min to LOC Track Mode

64% outside of coverage!Analyzed “Shortest time to LOC track after crossing 35° line”

All flights have 120 seconds or moreFrom 15° line, 71% still have 120+ sec120 sec before LOC TRACK used as time window criteria

13 ISPA 2008

High Workload Approaches

Workload highest if pilot busy keeping up with ATC instructions (short-cuts)High energy state / steep descent

EMS: “Excess Energy to Bleed” Warnings

3° G/S (data check)

14 ISPA 2008

Taskload Analysis

Difference between Workload (mental) and Taskload (actions)EMS can only measure Taskload – used as indirect measure of workloadSplit into Pilot Flying PF / Pilot Non Flying PNF

Each action triggers an event of 5 seconds (takes into account some initiation and feedback time)

Function of Autopilot (A320 peculiarity)Not possible to fully extend speedbrakes with AP ONPNF takes over all FMS / AP manipulations while PF hand-fliesPF looks idle due to absence of events, but is fully tasked

Analyzed Flights with Fast Intercept Geometry & High Workload

15 ISPA 2008

Taskplot Example

High Workload, Autopilot Off exampleGear, flaps, speed brakes, VHF voice, FMS & Autopilot manipulations, etc…ASSUMPTION: If 30 sec task free window within 120 to LOC track, IDENT can be done

30 sec task free

window

120 sec to LOC TRACK

16 ISPA 2008

Results: Distribution of Task Free Windows

Around 70% of flights meet the window citeria for either PF or PNF & AP on/off casesAs either pilot can do IDENT, cumulative probability is high

Remaining most difficult cases reviewed by expert pilotIDENT considered possible even in worst-worst casesBut overall argument is a stretch!

(30 sec bins)

17 ISPA 2008

IDENT & Workload Lesson Summary

Pilots need IDENT at check for approach near FL100If pilots had to wait with IDENT until being inside, coverage requirements would be challenging to justify (would need to be larger)

Not possible to remove this technical over-performance once pilots have gotten used to it!

Even if wrong IDENT is possible (co-channel / spectrum)

High Workload caused by high energy state / ATC shortcutConversely, pilots aiming for low workload come in low on purpose (long-haul)

18 ISPA 2008

Supporting Operational Studies

Check for Approach & IDENTWhat the pilots do with ILS outside of coverage…IDENT inside coverage and workload

ILS Intercept and CaptureFalse capture issue and coverage

ILS Tracking and Dynamic Multipath

19 ISPA 2008

Intercept Turn Initiation

Course guidance is on the order of ±2° (FSD, function of runway length)DDM’s typically keep increasing up to ±4°

This system margin is intended to give installation flexibility to the ANSPReliable intercept turn initiation without overshoot needs at least ±5°

More in strong tailwind situationsHas resulted in some avionics that make use of non-standardized regionNatural response to operator expectation, but leads to interoperability problems

Utilized by AFCS (up to ±4°)

Standardized Proportional RegionProblematic Scallops

20 ISPA 2008

15°

False Capture Mitigation (Partial)

Good practice to arm for LOC intercept as late as possible

Usually done near 5°Conversely, late arming is natural result of receiving ATC intercept clearance

21 ISPA 2008

Localizer Intercept Procedure

LOC Intercept GS Intercept

IF FAPIntermediate

Segment “Final Segment”

THR

RWY Length + LOC OffsetInitial Segment

Interaction of PANS-OPS, Flight Ops & Vectoring…

22 ISPA 2008

ILS Capture Points

Majority of Glide Path captures and significant portion of LOC capture occur well outside of formal coverageOK because centerline coverage is usually best

25 NM (LOC)

10 NM (GP)

23 ISPA 2008

Intercept and Capture Lesson Summary

Operational need for deviation guidance enabling reliable ILS intercept turn initiation is at least 5°

Pilot crews have gotten used to finding alternate meansRNAV or Airport VOR

Current interoperability issues (false capture) due to operational need

Operational use of ILS has evolvedAdvances in aircraft and associated aerodynamics, as well as ATC factors (vectoring practices) contribute to need for centerline guidance well outside of coverage ranges

Advances in Continuous Descent Approach (CDA) ImplementationWill increase desire for far out captures (LOC and especially GP)

24 ISPA 2008

Supporting Operational Studies

Check for Approach & IDENTWhat the pilots do with ILS outside of coverage…IDENT inside coverage and workload

ILS Intercept and CaptureFalse capture issue and coverage

ILS Tracking and Dynamic MultipathWhat level of deviations & disturbances is operationally relevant?

25 ISPA 2008

ILS Tracking - in Presence of Multipath

Supports ongoing Critical & Sensitive Area (CSA) work

Avionics have implemented additional filters to limit effect of transients (due to maneuvers on airport)

One OEM using 7 second window

Loss of LOC tracking is very rare (go-around or other maneuver)Prior to FAP ( >> 7 NM to THR), significant maneuvers and recapture possible (operational reasons - flock of birds, etc. but extremely rare)Roughly 1 or 2 in 300 go-around’s can be attributed to ILS

Transient deviation disturbance usually tolerable when visualGo-around likely in Instrument Met Conditions

26 ISPA 2008

ILS Deviations – How much is significant?

Half Dot (1/4 FSD, 37.5µA)

Half dot deviation maxima is quite significantNear 30µA SARPs tolerancereasonable limit even if derived to limit displacement at DH/DA

Events occur from 7 NM inward (5 sec example below)

27 ISPA 2008

Tracking & Overall Coverage Lesson Summary

Once aircraft is established on ILS, landing success highly likelyILS is a “minority cause” for rare approach interruptions and go-around’s

Many ILS operations each day pilots handle these cases regularlyAirline ¼ dot go-around criteria reflects operational experience

Pilots are NOT expected to know about coverage volumes and associated implications

ANSP notion of responsibility limit not relevant to pilotTo pilot, avionics are to indicate any limitations

A signal will be used wherever it is receivedANSP also has a desire to maximize utility of investmentPilots aware of limitations mostly from operational experienceIt can also be viewed as a detriment to safety to not use an available signal

28 ISPA 2008

Conclusions

ILS is considered a standard of operational suitabilityDoes not mean that there isn’t room for improvementNew landing systems should try to alleviate current limitations (esp. coverage)

“ILS look-alike” includes the non-standardized features…!

All such development requires an open dialogue between technical and operational communities

EMS Essential to Provide Quantitative, Data-Driven AnalysisLarge sample statistics global credibilityProvides filtering to enable expert analysis of small sample data

29 ISPA 2008

Questions?